This vaguely rings a bell, though from a long time ago. We had our id field using the "lowercase" type in Solr, and that broke/changed somewhere in the 4.x series (we are on 4.8.1 now and it doesn't work there), so we have to revert to a simple "string" type instead. I know you have a very different use case, but I don't think its anything to do with CDCR or 6.x, I think its a "problem" in the 4.x series. You might want to check the 4.x release notes, and/or try upgrading to 4.10.4 (the latest in the 4.x series) just to see what the behavior is there, I think it changed somewhere around 4.4 or 4.6...
But I'm talking probably 2-3 years ago, so my memory is hazy on this. On 9 August 2016 at 08:51, bharath.mvkumar <bharath.mvku...@gmail.com> wrote: > Hi All, > > I have an issue with cross data center replication, when we delete the > document by id from the main site. The target site document is not deleted. > I have the id field which is a unique field for my schema which is > configured as "long". > > If i change the type to "string" it works fine. Is there any issue using > long. Because we migrated from 4.4 to 6.1, and we had the id field as long. > Can you please help me with this. Really appreciate your help. > > I see the below error on the target site:- > > o.a.s.h.RequestHandlerBase org.apache.solr.common.SolrException: Invalid > Number: > at org.apache.solr.schema.TrieField.readableToIndexed( > TrieField.java:537) > at > org.apache.solr.update.DeleteUpdateCommand.getIndexedId( > DeleteUpdateCommand.java:65) > at > org.apache.solr.update.processor.DistributedUpdateProcessor.versionDelete( > DistributedUpdateProcessor.java:1495) > at > org.apache.solr.update.processor.CdcrUpdateProcessor.versionDelete( > CdcrUpdateProcessor.java:85) > > Thanks, > Bharath Kumar > > > > -- > View this message in context: http://lucene.472066.n3. > nabble.com/Unique-key-field-type-in-solr-6-1-schema-tp4290895.html > Sent from the Solr - User mailing list archive at Nabble.com. >
